Definition and value of everyday interactions

Early caregiving experiences set the foundation for developmental processes that unfold across the lifespan (Bowlby, Reference Bowlby1982). According to attachment theory and the organizational perspective on development, everyday interactions between a caregiver and child have cascading effects on attachment and other developmental processes (Doyle & Cicchetti, Reference Doyle and Cicchetti2017). Specifically, everyday interactions are characterized as social exchanges between the caregiver and child, and include the frequency, duration, and quality of these exchanges. Interactions occur over short bursts of time (e.g., microlevel interactions occurring within seconds) or can be characterized as long-standing patterns that accumulate over years (macrolevel) (Ram & Gerstorf, Reference Ram and Gerstorf2009). Assessing everyday interactions enables researchers to gain a snapshot of microlevel interactions that, over time, may represent the longer-term macro view of the relationship.

Laboratory assessments of caregiver–child interactions are the field standard, given the complexity of naturalistic environments and value of standardizing assessment methods. In laboratory assessments, the observations tend to be relatively brief in duration (∼5-45 minutes) and either feature behavioral instructions for the caregiver to act in a prespecified manner (e.g., be non-responsive to the infant during episodes of the Still Face Paradigm; Tronick et al., Reference Tronick, Als, Adamson, Wise and Brazelton1979) or to play with their child as they typically would (e.g., during “free play” interactions where a set of toys are made available to a caregiver–child pair; see Baldwin et al., Reference Baldwin, Baldwin and Cole1982). One value of these highly controlled settings is that variations in caregiving interactions are observed free from other external distractions that can influence the engagement of the participants (e.g., preparing meals, phones and tablets, other children). In addition, laboratory tasks allow researchers to conduct experiments that test specific research questions (e.g., how do caregiver–child dyads navigate more challenging tasks, including observing how caregivers instruct and support children to independently “clean up” appealing toys [see Kochanska & Aksan, Reference Kochanska and Aksan1995], or how the dyad responds to a brief separation [see Ainsworth et al., Reference Ainsworth, Blehar, Waters and Wall1978; Crowell & Feldman, Reference Crowell and Feldman1988]). Highly controlled lab paradigms have provided profound insight into developmental processes, including identifying the quality of interactions that precede the development of secure attachments (e.g., Braungart-Rieker et al., Reference Braungart-Rieker, Zentall, Lickenbrock, Ekas, Oshio and Planalp2014).

Knowledge gaps

As with any method, there are limitations to what can be gained via laboratory observations. First, behaviors observed in the laboratory capture the child’s experience with that caregiver under very specific confines. The caregiver is often acutely aware of being observed either in real-time by members of the research team or through cameras recording the interactions. Further, the removal of the necessary household tasks, phones, and other children that may otherwise pull the caregiver’s attention are likely meaningful moderators of the behaviors observed in the laboratory (e.g., McDaniel, Reference McDaniel2019), which may constitute the caregiver’s “best case scenario” of behavior. In addition, the observation is brief, and the degree to which the quality of behavior can be maintained over time is not typically tested, as social interactions are taxing energetically. In a study of mothers with and without sleep disturbances, those with poor sleep continuity demonstrated a notable decline in sensitivity across a 10-minute free play with their 18-week-old infants (King et al., Reference King, Rangel, Simpson, Tikotzky and Manber2020). Further, this higher effort during observations is evident from comparisons of caregiver behavior in these settings compared to daylong observations. Caregivers are likely to speak to their infants throughout the entire duration of a laboratory task, however 20–30% of everyday activities that infants engage in do not contain any adult speech (Cristia et al., Reference Cristia, Lavechin, Scaff, Soderstrom, Rowland, Räsänen, Bunce and Bergelson2021; Tamis-LeMonda et al., Reference Tamis-LeMonda, Kuchirko, Luo, Escobar and Bornstein2017). Taken together, although we know that laboratory assessments can provide insight into developmental processes, it is still unclear the degree to which the behaviors observed in the lab map on to what the child experiences in terms of caregiving in their everyday life. Although there is some recent work examining the role of caregiving “in the wild” in developmental processes (e.g., Madden-Rusnak et al., Reference Madden-Rusnak, Micheletti, Dominguez and de Barbaro2023; de Barbaro et al., Reference de Barbaro, Khante, Maier and Goodman2023), we know little about the level (and impact) of variation in caregiving (e.g., at home vs. at daycare; during meals vs. at bedtime; with mom vs. dad vs. when both are present; and in other combinations) on children’s functioning.

Recommendations for next steps

Infants' daily experiences encompass a wide array of interactions with various objects, individuals, and activities across settings. The deployment of tools to document these experiences can significantly enhance our understanding of how actions, proximity, tactile engagements, and exposure to varying environments influence child development (de Barbaro & Fausey, Reference de Barbaro and Fausey2022). Researchers may pursue these questions by considering first-person recordings in naturalistic contexts. Advances in wearable technology, such as the Looxcie 2 camera, which fits over one of the adult wearer’s ears – in this case, the caregiver – like a pair of glasses (e.g., Jayaraman et al., Reference Jayaraman, Fausey and Smith2015; Teti et al., Reference Teti, Kim, Mayer and Countermine2010), offer innovative ways to record infants' interactions without the discomfort associated with traditional head-mounted devices. Audio recording tools like the Language Environment Analysis system (Xu et al., Reference Xu, Yapanel and Gray2009), in which the device is typically housed in a wearable vest, gathers audio data from the infant’s environment, including speech and background sounds. Their “black box” algorithms produce estimates of adult words, conversation turns, and child vocalizations, offering the potential to use the data to obtain information such as which conversational partner initiated a conversation, and whether a greater number of adult- vs. infant-initiated conversations better explain children’s later expressive and receptive language (Salo, King et al., Reference Salo, King, Gotlib and Humphreys2022). Open-source tools that allow investigators to explore and manipulate the algorithms used are ideal for promoting collaboration and allowing for innovations that improve measurement.

Advances in wearable technologies also allow for new areas of inquiry. Specifically, physical proximity is an important aspect of the caregiver–child relationship (see Barnett et al., Reference Barnett, Hansen, Bailes and Humphreys2022), and can be recorded via touch and proximal distance recordings (Brzozowska et al., Reference Brzozowska, Longo, Mareschal, Wiesemann and Gliga2021; Salo, Pannuto et al., Reference Salo, Pannuto, Hedgecock, Biri, Russo, Piersiak and Humphreys2022). Such recordings provide valuable data on infants’ movement, spatial relations, and tactile interactions. Our team developed the TotTag (run using the SociTrack platform [Biri et al., Reference Biri, Jackson, Thiele, Dutta and Pannuto2020]), which employs time-of-flight technology to record continuous distance measurements between all individuals in a network. Importantly, this means that each pairwise relationship (not only mother–infant, for example) is captured. For example, in the case of TotTags worn by all members of a family of five, proximity information is produced for ten pairwise relationships. Data analysis options include examining combinations of close contact and considering the family unit’s patterns of proximity as a network. Critical aspects of caregiver–child relationship dynamics, including the duration of close proximity, frequency of caregiver–child “check-ins,” and which actor initiates interactions using the on-board accelerometer, may be important for evaluating differences between families, changes in behaviors across development, risk for developmental delays, and assessments of intervention success. Moreover, it can detect periods when a caregiver is out of the child’s range, providing insights into the physical (with implications for emotional) availability and responsiveness of the caregiver (for an example using another method to assess availability, see King et al., Reference King, Querdasi, Humphreys and Gotlib2021). Being able to include these details about everyday interactions may be essential for better understanding how caregiver physical contact influences whether and when children meet various developmental goals (e.g., establishing independence, providing children with feelings of security) and what may be the most effective behaviors to target for intervention.

Research to date has begun to exemplify how understanding physical proximity can support targeted intervention. For instance, studies highlight the potential impact of physical touch on the formation of secure attachment (Williams & Turner, Reference Williams and Turner2020) and demonstrate how an infant’s physical location within their environment can influence their exposure to language (Malachowski et al., Reference Malachowski, Salo, Needham and Humphreys2023; Mireault et al., Reference Mireault, Rainville and Laughlin2018) – all of which are vital to their developmental trajectory. This is only the tip of the iceberg in terms of important areas for research into everyday interactions, however. For example, research examining emotion socialization, or the ways in which caregivers respond to children’s emotions, which subsequently teach children how to understand and regulate their own emotions (Eisenberg et al., Reference Eisenberg, Cumberland and Spinrad1998; Eisenberg, Reference Eisenberg2020), has suggested that supportive emotional responses from caregivers (e.g., emotion labeling) have been associated with the development of self-regulation (P. Tan et al., Reference Tan, Oppenheimer, Ladouceur, Butterfield and Silk2020), sympathy for others (Curtis et al., Reference Curtis, Zhou and Tao2020), lower levels of behavior problems (Perry et al., Reference Perry, Dollar, Calkins, Keane and Shanahan2020; Zhang et al., Reference Zhang, Lee, White and Qiu2020), and greater well-being (Brown & Fredrickson, Reference Brown and Fredrickson2021). However, the degree to which caregiver emotion socialization measured in the laboratory or via questionnaires extends to everyday experiences – and the degree to which caregiver physical proximity is a prerequisite for these experiences – is unknown.

Pursuit of these questions will be enhanced by the use of better software and analytic approaches now available for the rich data discussed here. Machine learning and computer vision technologies are revolutionizing the analysis of video data, enabling researchers to rapidly detect and analyze faces, objects, and movements within video frames (e.g., Bambach et al., Reference Bambach, Crandall, Smith and Yu2018). These advancements significantly decrease the time and effort needed to analyze extensive video recordings. As a result, video coding, known for its time and labor intensity, becomes more accessible to researchers who lack the resources for hiring and training coders, as well as those seeking to access large volumes of recording. Moreover, advancements in the widespread availability and ease of ecological momentary assessment (EMA) provide an opportunity to advance work pertaining to everyday interactions and capturing caregiving in moments of heightened interest (e.g., during emotionally charged situations). These approaches would support further understanding of lingering questions surrounding in-the-moment caregiving. For example, in the case of emotion socialization, EMA can be applied to obtain information in shortened timescales (e.g., children’s emotions during the last 2 hours compared to longer timeframes specified in questionnaires). This would provide more accurate insight into how caregivers’ own emotions and behaviors in response to their children’s different affective states impact caregiving quality, potentially identifying targets to help support families in more challenging moments.

Compared to traditional laboratory observations, advances in statistical approaches and software, proximal distance recordings, and wearable technologies (i.e., camera and audio recorders) enable a more comprehensive and precise portrayal of early experiences that occur daily and in naturalistic contexts. Bridging the gap between laboratory observations and real-world experiences better captures the subtleties of daily life that contribute to infant and child development. Specifically, with these tools researchers can more accurately measure the quality and impact of caregiver–child interactions in broad and specific naturalistic contexts, paving the way for interventions that support healthy developmental processes in everyday life.

Definitions and value of caregiver social cognition

Knowledge gaps

Although promising, there are several gaps in our current understanding of caregivers’ social cognitions. First, despite theoretical support for the idea that social cognitions unfold over time with caregiver and child development, empirical studies focusing on the reciprocal development between caregivers’ cognitions and children’s functioning and development remain very limited. As children advance through different stages of development, the synchrony in the mental engagement between caregiver and child – this “meeting of the minds” – is likely to transform (Sharp & Fonagy, Reference Sharp and Fonagy2008). Thus, it is reasonable to propose that as children acquire new abilities, the nature and elaborateness of the caregiver’s social cognitions would plausibly adapt in response to these developmental progressions. Yet which changes or cues signal to caregivers that the child is capable of more mature considerations is not clear.

Congruently, recent studies have focused on the progression of caregivers’ mind-mindedness (Bigelow et al., Reference Bigelow, Power and Dadgar2023; Foley et al., Reference Foley, Devine and Hughes2023). These studies have found a degree of consistency, as well as some variability, in caregivers’ comments about their child’s mental world from the prenatal and newborn phases to three to four months postpartum. Specifically, one study found an uptick in mothers’ attuned mind-minded remarks toward their infants over the initial three months, while non-attuned comments remained unchanged (Bigelow et al., Reference Bigelow, Power and Dadgar2023). Despite these findings, the field predominantly treats reflective functioning as a stable characteristic. Whether social cognitions are truly fixed or subject to change in response to a child’s developmental progress remains an area poised for further exploration. For instance, caregivers communicate differently with infants as a function of whether they have achieved certain motor milestones (Schneider & Iverson, Reference Schneider and Iverson2022; West & Iverson, Reference West and Iverson2021). The pattern, nature, and individual differences in caregivers’ “updating” their models is an important gap in our understanding of the bidirectional relationship between caregiver cognitions and children’s evolving capabilities.

Second, in theoretical models, social cognitions have been understood as dynamic, capable of fluctuating with changes in context (Fonagy et al., Reference Fonagy, Steele, Steele, Moran and Higgitt1991). Yet, in practice, our study designs and measurements typically capture the relationship at a set point in development, or, if longitudinal, represent either snapshots of time assessed across development or capture more global tendencies. What remains to be considered are specific timescales of the ideal assessments, and further, whether there are some contexts in which mental state reflection may be more or less important (for an example of the importance of caregiver sensitivity varying based on the affective context, see Leerkes et al., Reference Leerkes, Blankson and O’Brien2009). Evidence from research on empathy further suggests effects of psychological symptoms specifically on parental cognitions. Depressive symptoms were linked to lower levels of empathy toward one’s own child over and above the association between depression and dispositional, or general, empathy (Salo et al., Reference Salo, Schunck and Humphreys2020). Further, given that the ability and tendency to consider a child’s mental states may be influenced by the caregiver’s own temperamental or developmental attributes (Sharp & Fonagy, Reference Sharp and Fonagy2008), understanding how caregivers’ mental and emotional states change as a function of context, including the effects of acute stress as well as functionally impairing depression, for example, is critical for future research.

Third, the measurement of such complex constructs represents a sizeable challenge for research, and current scoring schemes and self-reported assessments have limitations. For parental reflective functioning specifically, existing coding systems produce a single, unidimensional score, which may not fully capture the nuances of caregivers’ abilities to reflect upon the thoughts and feelings of both themselves and their children (Garon-Bissonnette et al., Reference Garon-Bissonnette, Dubois-Comtois, St-Laurent and Berthelot2023; Sleed et al., Reference Sleed, Isosävi and Fonagy2021). This scoring limitation can obscure significant variances in impairments in reflective functioning among caregivers, who might exhibit comparable scores but possess distinct mental content and conceptualizations of their child. For instance, a caregiver who distances themself from thinking in mental states terms would receive a low score of reflective functioning, and at present the same score could also be assigned to a caregiver who attempts reflection but provides confused, diffuse, or inconsistent responses (Garon-Bissonnette et al., Reference Garon-Bissonnette, Dubois-Comtois, St-Laurent and Berthelot2023). Similarly, as parental reflective functioning considers both reflection regarding the child as well as self as a parent, differences in caregiver- versus child-focused reflective functions specifically may emerge (see Smaling et al., Reference Smaling, Huijbregts, van der Heijden, van Goozen and Swaab2016). These qualitative differences may be critical in informing intervention targets for a given caregiver–child pair (Sleed et al., Reference Sleed, Isosävi and Fonagy2021) and relatedly in providing more personalized interventions.

Finally, caregivers’ social cognitions and representations may promote responsiveness to caregiving interventions or evolve as a result of interventions. For instance, one recent study indicated improvement in parental reflective functioning after a short-term intervention aimed at bolstering attachment security, improvements that explained significant variance in positive caregiving behaviors (Dexter & Wong, Reference Dexter and Wong2023). Relatedly, a separate study indicated that pre- to post-intervention changes in caregivers’ beliefs about their importance for child development were linked to more positive dyadic interactions, higher learning, and socioemotional skills in children (List et al., Reference List, Pernaudet and Suskind2021). While this preliminary research is promising, the impact of this work would be further enhanced with pre-registered replications and extensions of this work.

Recommendations for next steps

Current measurement techniques might not adequately differentiate aspects of caregivers’ reflective capacities, potentially leading to uniform treatment approaches when more personalized strategies may be advantageous. Recent advances have suggested adaptations to existing coding schemes that have the ability to tap into the representational and perceptual heterogeneity of caregiver cognitions (Garon-Bissonnette et al., Reference Garon-Bissonnette, Dubois-Comtois, St-Laurent and Berthelot2023; Sleed et al., Reference Sleed, Isosävi and Fonagy2021). These approaches pave the way for tailored methods to support caregiver–child relationships. Specific forms of failures to mentalize were shown to have different determinants in the caregiver’s own developmental history; differences otherwise masked using the usual approach to score reflective functioning (Garon-Bissonnette et al., Reference Garon-Bissonnette, Dubois-Comtois, St-Laurent and Berthelot2023). Similarly, the different representational contents (classified as hostile, helpless, or idealized) of caregivers’ discourse when asked to reflect on their children differentially correlated with the attunement and mutuality between caregiver and child. For instance, hostile and idealized representations, characterized by lower reflective functioning, were associated with poorer caregiver–child attunement, and helpless representations were not strongly associated with measures of caregiver–child relationship quality during a brief laboratory assessment (Sleed et al., Reference Sleed, Isosävi and Fonagy2021). Such “splitting” (e.g., rather than lumping) approaches may better yield information necessary to understand how child-related cognitions develop as well as which dimensions matter most for the caregiving relationship, and in turn the child’s cognitive, social, and emotional development. Future investigations into these complex dimensions of caregiver social cognitions hold promise for the development of more sophisticated and effective assessment tools (King, Salo, et al., Reference King, Salo, Kujawa and Humphreys2021).

Similarly, narrowing the focus from a caregiver’s general- versus child-specific ability for social cognitions (e.g., reflective functioning and parental reflective functioning, dispositional empathy and empathy towards child specifically), and perhaps even in specific moments within the relationship, merits attention in future research (see Salo et al., Reference Salo, Schunck and Humphreys2020; Smaling et al., Reference Smaling, Huijbregts, van der Heijden, van Goozen and Swaab2016). Research from our team demonstrated important variation in caregiving behaviors even within the relationship with a given child (Bailes et al., Reference Bailes, Ennis, Lempres, Cole and Humphreys2023). This study found that caregivers varied in the degree they felt anger and frustration in response to their child’s emotional displays, which were also dependent on whether the emotional expression was considered to be justified or unjustified. Adopting a more granular lens – one that focuses on a given child, a specific relationship, a moment in time, and subtleties of processes, rather than adopting a general, or even caregiver context–specific approach – may be critical in understanding the interrelatedness and complexity of social cognitive and emotional processes.

As previously noted, caregiver–child dynamics are constantly evolving over time, both as the relational partners learn about one another and as the relationship changes in response to growing competencies for both caregiver and child, especially due to the dramatic physical and mental changes occurring within the developing child. Exploring caregivers’ understanding of their child’s mental capacity offers valuable insights into how this awareness may shift with the child’s development. The concept of “mind perception” from cognitive science relates to our judgments about others’ mental capabilities and has been shown to correlate with a greater liking and valuing of that entity (Gray et al., Reference Gray, Gray and Wegner2007). Making more accurate judgments about children’s capabilities, such as understanding their intentions, is crucial for caregivers. It guides the support they provide in areas like motor skills, language, and cognitive tasks. For example, infant communicative behaviors perceived as intentional by caregivers were shown to be more likely to elicit a contingent behavioral response from caregivers than communicative behaviors perceived as non-intentional (Donnellan et al., Reference Donnellan, Bannard, McGillion, Slocombe and Matthews2020). Furthermore, interpreting behaviors more generously – for instance, considering non-blame-oriented reasons for infant night wakings or toddler temper tantrums – can prevent harsh responses or abuse (Bugental et al., Reference Bugental, Ellerson, Lin, Rainey, Kokotovic and O’Hara2002). This connection between a caregiver’s perceptions of their child’s evolving capacities and the caregiver’s responsive and constructive behaviors has been posited to influence the quality of the caregiving relationship, enhancing reflective functioning. Our research group has assessed mind perception in adults as they consider the minds of infants and young children throughout development (Weisman et al., Reference Weisman, King and Humphreys2024) and among caregivers (Salo et al., Reference Salo, Hare, Letterie, King, Weisman and Humphreys2024). We found estimates of children’s mental capacities increased non-linearly as a function of chronological age, with greater changes observed in infancy relative to the preschool-age years. Further, higher levels of estimated capacities were associated with higher levels of parental reflective functioning and more positive caregiving behaviors in parents of preschool-age children (Salo et al., Reference Salo, Hare, Letterie, King, Weisman and Humphreys2024).

Caregivers also develop in their new role, and experience in caregiving may further influence their cognitions regarding their child. For instance, pregnant individuals differ in their beliefs about children’s needs based on whether or not they have other children (Mascheroni et al., Reference Mascheroni, Grassi, Bonanomi, Sperotto, Deeg, Hung, Xia, Ionio, Kit‐fong Au and Gattis2022). Specifically, multiparous individuals prioritize attunement, while primiparous individuals place greater value on structure. Within-person, longitudinal research (e.g., studying the same adult across years to follow their journey as a parent with multiple children) is needed to evaluate how to disentangle caregiving experience with child age/developmental stage in their influence on a caregiver’s cognitions and behaviors.

In sum, much evidence suggests that social cognitions are central contributors to behaviors and development and are important modifiable targets for caregiver–child interventions. Still, however, many gaps in the literature remain to be addressed. Future research would benefit from exploring caregiver social cognitions regarding their child using assessments and coding schemes that capture different dimensions and distinguish between general, child-specific, and even situation-specific abilities and tendencies. In addition, inclusion of longitudinal designs that permit exploration of how perceptions and behaviors change in concert across the first years of life, including how they relate to contextual factors, is an important area for future research.

Definition and value of considering context for caregiving relationships

Developmental psychopathology, as mentioned above, has long considered children’s contexts in influencing their outcomes (Cicchetti & Aber, Reference Cicchetti and Aber1998). In keeping with this framework, we consider how the broader context in which a caregiver lives impacts their behavior toward, cognitions about, and ultimately, their relationship with their child (bridging from the macro- to the microsystem; Cicchetti & Aber, Reference Cicchetti and Aber1998). Further, caregiver context specifically impacts the real-life everyday interactions the child has in terms of physical access to their caregivers and the emotional availability of the caregiver during those opportunities for interactions. Notably, we specify caregiver context here, rather than family context, to specify factors that influence the child through their impact on caregiver functioning and availability. While caregiving relationships are influenced by a wide range of environmental, societal, cultural, and personal factors, here we discuss caregiver’s (1) relationships with their romantic partners, (2) social support networks, (3) policies regarding post-birth (paid and unpaid) leave, and (4) culture. These factors represent only a few areas to consider within the broader context that impacts caregivers and their relationship with their children. While each is discussed independently, these factors are often related and have bidirectional impacts. For example, cultural norms about caregiver roles may influence expectations regarding dividing childcare between partners (Craig & Mullan, Reference Craig and Mullan2010; Fernández & Fogli, Reference Fernández and Fogli2009); partner relationships provide an important source of social support (Goldberg & Carlson, Reference Goldberg and Carlson2014; McRae et al., Reference McRae, Overall, Henderson, Low and Chang2021); and some configurations of parental leave policies may reinforce expectations that birthing parents are most responsible for a child’s caregiving (Schober, Reference Schober2011).

Knowledge gaps